Considerable interest has been generated by a program of study of Human Genome Diversity aimed at collecting world-wide sample of representative human populations to be made available for systematic analysis by molecular geneticists. In addition to preparing the sample of DNAs representing world variation it seems also necessary and timely, however, to collect a battery of genetic markers worth a systematic and coordinated investigation. This is a non trivial problem given the number of the individuals to be tested, the need to select a large balanced set of new markers and representing the whole genome, useful for their evolutionary and health relevance. A main goal of the present program project is to help in this task. There already exists a minicollection of cell lines of populations representing the whole world which can be employed for a pilot study. This collection can produce results valuable per se as well as for the organization of the analysis of the Human Genome Diversity material expected in the future. In principle, the most thorough way of studying genetic variation is at the level of DNA sequences approach already useful for two systems: the control region of mitochondrial DNA, and several HLA genes. They are both characterized by very high variation. The search of highly variable sequences is a valuable aim in many ways: some have already proved to be quite useful also for evolutionary analysis, like microsatellites, and demand further study. Certain chromosomes, in particular the Y chromosome, are in need of special, widespread attention. New molecular methods now operational or nearly so offer potential breakthroughs and must be compared with the old ones. Interpreting sequence variation demands new mathematical methods, of which there exist already a few examples, often in need of refinement and further testing with real data. It is important to establish close collaboration among experimenters and mathematicians, which has been essential in physics but is not common in biology. The present program project is structured so as to offer continuous feedback between theory and experimentation.